GitHub Repository

Long-term memory for LLMs. MCP server backed by hybrid search in a single SQLite file.

48 starsPython

Rekal – Long-term memory for LLMs in a single SQLite file

Name: Rekal – Long-term memory for LLMs in a single SQLite file
Availability: InStock
Author: jeeybee

by jeeybee·Apr 12, 2026·9 points·10 comments

Visit Project View on HN

AI Analysis

●●SolidSolve My ProblemSlick

Single-file SQLite memory for LLMs simplifies complex vector DBs for local workflows.

Strengths

•Hybrid search blends BM25, vectors, and recency decay using local embeddings only.
•Conflict detection and superseding tools prevent stale knowledge from poisoning context.
•Claude Code skills automate hygiene and saving without manual prompt engineering.

Weaknesses

•Requires Python 3.14+, which limits adoption on systems stuck to 3.11.
•Single-file architecture might struggle with concurrency if multiple agents write simultaneously.

Post Description

I got tired of repeating myself to my LLM every session. rekal is an MCP server that stores memories in SQLite and retrieves them with hybrid search (BM25 + vectors + recency decay). One file, local embeddings, no API keys.